The primary goal of this program is the continuing development of a probabilistic reference system for the human brain as an important neuroinformatics tool for use by the neuroscience community. In our original application we proposed to build the tools and determine the feasibility of this approach in the domain of structural neuroanatomy. Both goals have been achieved. In this renewal application, we propose to add functional (fMRI and PET) and microscopic structural (cyto- and chemoarchitectural) information about the brain to this system through the development of unique informatics tools and strategies. An important neuroscience outcome of this plan will be the opportunity to examine, for the first time, the stability, relationships and distribution of micro- and macroscopic structure to function in the human brain. This issue, although a major area of interest, has remained a vexing problem because of the difficulty in obtaining data sets of sufficient magnitude, diversity, number and organization to answer such questions. These goals will be achieved through an integrated program made up of six separate institutions (UCLA, Montreal Neurologic Institute, University of Texas at San Antonio, Stanford University, Albert Einstein College of Medicine and H. Heine University in Dusseldorf). Six specific aims are proposed: 1) Develop the neuroinformatics tools, algorithms and protocols to integrate functional information about the human brain, from fMRI and PET, through the identification of functional landmarks. 2) Establish and validate a Functional Reference Battery (FRB) of behavioral tasks that reliably produces functional landmarks across subjects and modalities 3) Extend the range, magnitude and attributes of structural data from both in vivo and post mortem sources; the latter to include cyto-and chemoarchitectural information for selected brain regions. 4) Develop the statistical and warping pools to analyze these data sets in a unified and formal multivariate environment. 5) Create a database structure, organized in four dimensions and tested in three, designed for query-by-content of large, multi-attribute data sets. 6) Beta test, at new sites beyond our consortium, those developed through the program, including independent evaluation of their commercial value. The use of a consortium structure where the distribution of labor can be separated into parallel, complementary tasks, executed by established leaders in the field, has been efficient, in both cost and time, as compared to isolated efforts at each site. It has also created a "real world" environment among participants such that differences in equipment, software and protocols actually reflect a microcosm of the larger neuroscience and neuroinformatics communities.